Our electric grid is the most massive machine that humans have ever built. The grid makes every part of our modern lives possible, but this massive machine isn’t perfect. For past 100 years, the electricity grid relied on a single power source and did not provide detailed information on the usage data. This made electricity difficult to manage resulting in massive power outages. The current centrally controlled power grid is a large interconnected network that delivers electricity from suppliers to consumers. Real-time, bi-directional communication is imperative to establishing optimal and balanced energy management in the grid thereby eradicating disastrous blackouts that have been haunting the nation more often than acceptable in the recent past. The smart grid is the country’s electric grid for the 21st century. The smart grid is conceived to have the ability to adopt technologies covering the areas of sensing, wireless connectivity, pervasive computing and adaptive control to significantly improve the efficiency. The amount of data handled by utilities in the U.S today is around 75000 TB. This tremendous amount of data transaction necessitates exploration of new spectrum bands and better spectrum management beyond what is available today. In this thesis, we formulate a smart real time power update algorithm with minimal power overhead. We employ an efficient power saving channel access mechanism using IEEE 802.11af TV white space (TVWS) spectrum. Since most of the other unlicensed bands (such as the ISM bands) are already used heavily, TVWS opens up a new and relatively untapped resource. We implement a MAC layer scheduling algorithm for wireless Community Area Network (CAN) system to reduce collision rate and transmission delay of smart meters to a utility base station in a neighborhood. Through our detailed simulation study, we show that the proposed scheme offers intelligent real time power updates to power facilities with minimal power overhead.